Associations of Insulin and IGFBP-3 with Lung Cancer Susceptibility in Current Smokers

Abstract

Background: The epidermal growth factor receptor (EGFR) signaling network is involved in lung carcinogenesis. This study examined whether ligands that activate or suppress the EGFR signaling network were associated with lung cancer risk in ever smokers.

Methods: A nested case-control study within the Women's Health Initiative assessed baseline plasma levels of insulin, insulin-like growth factor (IGF)-1, insulin-like growth factor binding protein (IGFBP)-3, interleukin (IL)-6, hepatocyte growth factor (HGF), and nerve growth factor (NGF) in 1143 ever-smoking lung cancer cases and 1143 controls. Leptin was measured as an adiposity biomarker. Conditional logistic regression was used in data analyses.

Results: Leptin was inversely associated with lung cancer risk (odds ratio [ORcontinuous] per Ln [pg/mL] = 0.85, 95% confidence interval [CI] = 0.74 to 0.98). After adjusting for adiposity and other risk factors, null associations were found for IL-6, HGF, and NGF. In current smokers, but not former smokers, high insulin levels were associated with increased lung cancer risk (OR for 4th quartile vs others [ORq4] = 2.06, 95% CI = 1.30 to 3.26) whereas IGFBP-3 had a linear inverse association (ORcontinuous per μg/mL = 0.64, 95% CI = 0.41 to 0.98). The insulin association was consistent across subgroups defined by body mass index and histological type, but the IGFBP-3 association was specific to small cell lung cancer. There was a modest positive association between IGF-1 and lung cancer risk in current smokers (ORq4 = 1.44, 95% CI = 0.90 to 2.29).

Conclusions: Independent of obesity, high insulin levels but reduced levels of IGFBP-3 were associated with increased lung cancer risk in current smokers.

Figure 1.

Adjusted odds ratios (ORs) and 95% confidence intervals (CIs) for the associations of lung cancer risk with insulin, IGFBP-3, and IGF-1 in current and former smokers. Odds ratios were obtained from one multivariable conditional logistic regression model with the three analytes (insulin, IGFBP-3, and IGF-1) analyzed in quartiles and their multiplicative interaction terms with smoking status, adjusting for age, smoking history (pack-years of smoking, age started smoking, and years of smoking cessation), lifestyle (body mass index [BMI], plasma level of leptin, servings of fruits per day, and total energy intake per day in kilocalories), and medical history (history of emphysema, white blood cell count, years between menarche and menopause, and use of diabetes treatment). Cutpoints for quartiles were ≤93.4, 93.5–130.2, 130.3–191.6, ≥191.7 (pg/mL) for insulin; ≤1.77, 1.78–2.18, 2.19–2.62, ≥2.63 (μg/mL) for IGFBP-3; ≤57.4, 57.5–71.8, 71.9–88.2, ≥88.3 (ng/mL) for IGF-1. All P values were two-sided. No. Ca/Co = Numbers of cases and controls.